1. Field of the Invention
The invention relates to a novel type of Gram-positive replicon and its use for the construction of recombinant vectors.
2. Discussion of the Background
The naturally occurring bacterial plasmids are usually endowed with a high degree of stability. If they are distributed randomly the theoretical probability (L) of producing a plasmid-free cell is given by the equation: L=(1/2).sup.2n (where n is the number of plasmid molecules per cell and it is supposed that 2n copies of the plasmid are present immediately prior to division). Thus, high copy number plasmids can be maintained in a stable manner following random distribution, whereas efficient mechanisms are necessary to prevent the loss of low copy number plasmids in the absence of selection pressure (Nordstrom and Austin, 1989 Annu. Rev. Genet. 23: 37-69).
In Gram-negative bacteria, several systems involving maintenance functions for the plasmids have been described for low copy number plasmids such as F, P1and R1. They comprise the hok/sok and ccd systems responsible for the post-segregational destruction of the cells which lose plasmids (Jaffeet al., 1985 J. Bacteriol. 163: 841-849; Gerdes et al., 1986 Mol. Microbiol. 4: 1807-1818) and true partition functions which ensure that each daughter cell receives a plasmid molecule (Austin and Nordstrom, 1990 Cell 60: 351-354). Furthermore, auxiliary functions which allow random distribution of the replicons have been described. Examples include the sites-specific recombination system of Col E1 (Summers and Sherrat, 1984 Cell 36: 1097-1103) and the par locus of pSC101 (Tucker et al., 1984 Cell 38: 191-201).
In Gram-positive bacteria, much less information is available concerning the functions required for the stable maintenance of the plasmids. In the plasmids with high or low copy numbers which replicate by a "rolling circle" mechanism through a single-stranded DNA intermediate (te Riele et al., 1986a EMBO J. 5: 631-637, 1986b Proc. Natl. Acad Sci. USA 83: 2541-2545; Gruss and Ehrlich, 1989 Microbiol. Rev. 53: 231-241), the stability is coupled to replication and not to a discrete function. In the case of various known plasmids pUB110, pTA1060, pMV158 (Bron et al., 1991b Plasmid 19: 231-241), and pLS11 (Chang et al., 1987 J. Bacteriol. 169: 3952-3962), the regions of stability are the "negative" origins controlling the conversion of the single-stranded DNA into double-stranded plasmid DNA. The replication of these plasmids is guaranteed by a protein encoded in the plasmid. These plasmids are frequently unstable from a structural and segregational point of view when they are used as cloning vectors, and this is probably due to their particular mode of replication.
A second class of Gram-positive replicons comprises the large plasmids with low copy numbers, pTB19, pIP404 and pAMbeta1 (Imanaka et al., 1986 Mol. Gen. Genet. 205: 90-96; Garneir and Cole, 1988 Plasmid 19: 151-160; Swinfield et al., 1990 Gene 87: 79-90). Their replication requires a protein encoded in the plasmid which does not share any homology with that of the plasmids whose replication involves a "rolling circle" mechanism. These plasmids do not accumulate single-stranded DNA during replication (Garnier and Cole, 1988 Plasmid 19: 151-160; Janniere et al., 1990 Gene 87: 53-61) and are structurally stable (Janniere et al., 1990 Gene 87: 53-61).
No information is available concerning the potential partition functions of these plasmids with low copy numbers However, these functions are essential for their maintenance in sporulating Gram-positive bacteria such as the Bacillus species. During the differentiation stage of these bacteria, the spore compartment represents only about one sixth of the cell, and the probability of producing a plasmid-free spore is: L=(5/6).sup.2n. A random distribution of the plasmids during sporulation may, consequently, lead to a serious loss of replicons (with n&lt;10) in the spore population.
The sporulating Gram-positive bacterium Bacillus thuringiensis is known for its capacity to produce insecticidal toxins during sporulation (Lereclus et al., 1989b Regulation of Procaryotic Development. American Society for Microbiology, Washington D.C., 255-276). These parasporal proteins (delta-endotoxins) are each classified as being either CryI, II III or IV as a function of their activity spectrum towards insect larvae (Hofte and Whiteley, 1989 Microbiol. Rev. 53: 242-255). Different cry genes coding for delta-endotoxins active against various orders of insects have been cloned, and transformation by electroporation now makes possible the analysis of their expression in their natural host and the construction of strains obtained by genetic engineering exhibiting improved properties for insect control. B. thuringiensis exhibits a complex series of resident plasmids in most of the strains examined (Gonzalez et al., 1981 Plasmid 5: 351-365; Lereclus et al., 1982 Mol. Gen. Genet. 186: 391-398; Kronstad et al., 1983 J. Bacteriol 154: 419-428). The quite widespread presence of these plasmids suggests very efficient stability and replication functions. Starting from that observation, cloning and characterization studies of small cryptic plasmids (Mahillon et al., 1988 Plasmid 19: 169-173; Mahillon and Seurinck, 1988 Nucleic Acids Res. 16: 11827-11828; McDowell et al., 1991 Plasmid 25: 113-120) and the isolation of the replication regions of large plasmids (Baum et al., 1990 Appl. Environ. Microbiol. 56: 3420-3428) and of two smaller plasmids pHT1000 and pHT1030 of 8.6 kb and 15 kb, respectively (Lereclus et al., FEMS Microbiol. Lett. 49: 417-422) have been accomplished. pHT1030 was studied on account of its high segregation stability in B. subtilis, and a 2.9 kb DNA fragment including the stability and replication function of pHT1030 was used to construct the shuttle vector pHT3101 (Lereclus et al., 1989a FEMS Microbiol. Lett. 60: 211-218; Debarbouilleet al., 1990 J. Bacteriol. 172: 3966-3973).